• State TI Thinks Resolved
  • Locked Locked
  • Replies 8 replies
  • Answers 3 answers
  • Subscribers 30 subscribers
  • Views 392 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

PGA460: PGA460 Specifications

Ichi
Prodigy 20 points
Part Number: PGA460
Other Parts Discussed in Thread: TUSS4470

Tool/software:

Hi

I have a question about the specifications of the ultrasonic transducer connected to the PGA460.

I am trying to measure distances of less than 5 cm using a transducer with a center frequency of 400 kHz.

I have changed the frequency setting of the burst pulse and been able to output a 400 kHz pulse.

I am trying changing the register settings for the threshold, BPF, and LPF.

This is a basic question, but can the PGA460 measure distances with a transducer like the one described above?

Thank you in advance.

  • 0
    TI__Mastermind 23045 points

    Hello Ichi,

    Thanks for posting to the sensors forum! 5cm is rather small for ultrasonic measurement that use only a single transducer. Using higher frequency helps reduce the blindzone caused on a single transducer since the wavelength is much smaller but other factors like voltage and amount of pulses transmitted can affect the minimum detectable distance with a single transducer. 

    For a system that can handle small distances I would recommend using a bistatic topology which requires the usage of two transducers to make the measurement. So one is strictly driving the ultrasonic pulse signal and the other one is purely receiving. Using this type of topology allows you detect a smaller range.

    Let me know if there are any other questions on this topic!

    Best,

    Isaac

  • 0 in reply to Isaac Lara
    Prodigy 20 points

    ISSAC, sorry for the delay in replying.
    I have tried changing various registers with single use transducers but cannot capture the echoes well.
    It is proving to be very difficult.
    Does TI have an AFE with a 400 kHz transducer with specifications that would capture a 300 µs echo?
    The VPWR for the electrical circuit is 12Vdc.
    I think TUSS4470 is suitable, what do you think?
    Best regards.

  • 0 in reply to Ichi
    TI__Mastermind 23045 points

    Hello Ichi,

    Thanks for the info. The TUSS44x0 line of devices are both compatible with 400kHz transducers, and the PGA460 is also capable of handling this frequency as well. As mentioned its not a spec that defines if we can measure that short of an echo pulse or not, this is a limitation of ultrasonic technology when using a single device. The most effective way to measure pulses this short especially at this large of a voltage would be to separate the pulse transmission from the pulse receiver. Meaning that you would need to separate circuits to achieve this measurement.

    The reason why you need multiple devices and transducers to achieve this is due to the fact that if you use a single transducer it becomes saturated during the transmission period so until the driving pulse decays then the device is blind to any return echoes that occur during that time. Thats why large voltages make it harder to detect shorter distances since the decay time will be longer.

    I hope this makes it more clear on how to achieve your 5cm measurements. As always let me know if there are any questions.

    Best,

    Isaac 

  • 0 in reply to Isaac Lara
    Prodigy 20 points

    Thank you for your quick response, Issac.

    As you say, the burst frequency is high and the voltage is high, so the period from the transmitted pulse to the return of the echo is short.

    As a result, the echo signal is buried in the area surrounding the burst pulse transmission.

    In fact, we have been using this ultrasonic sensor and an old analog circuit for sensing for decades.
    The ultrasonic sensor is placed in a metal capsule about 5 cm long, and the gas to be measured is passed through it to separate the gas type.
    I have attached the waveform observed using the current device and the waveform of the pulse transmitted from the PGA460.
    The current circuit amplifies the echo signal (80dB in two stages), passes it through the BPF and LPF, and rectifies it to obtain a DC signal waveform.
    We are able to extract the echo from the signal multiplied by 10,000.
    We are verifying whether we can achieve the same thing using TI's AFE, but do you think it is possible?

    Best regards.

    Ichi

    WAVES.pdf

  • 0 in reply to Ichi
    TI__Mastermind 23045 points

    Hello Ichi,

    Thanks for the added explanation here on your application and planned use of the sensor. Looking at the document you sent over definitely makes me believe you were looking at the wrong device for your application. If you are using an analog signal that has been amplified and filtered  like the one shown in your pdf then using something like TUSS4470 would be closer to the system you are currently using. It looks like your area of interest is still out of the way from the PGA460 burst, the difference with PGA460 is that everything is collected digitally because of the built ADC needed to digitally process the return data so you wont really be able to see the returned echo unless if you use one of the DSP output features on the device.

    It looks like based on your configurations you are definitely out of the blind zone/ decay period of the sound burst so detecting a signal at 300us should be possible, the PGA460 may just be a little more difficult since you need to program thresholds on the device in order for the device to report any results.

    Best,

    Isaac

  • 0 in reply to Isaac Lara
    Prodigy 20 points

    Hello ISSAC

    Thank you for your accurate advice.
    Can I ask you a question repeatedly?　　This is my last question.
    I also think that it is difficult to proceed with the evaluation with PGA460.
    However, could you please check the following parameters that I have set?

    ・EEPROM register
    ・Threshold

    The numerical settings of BPF and LPF are calculated and set from past forums here.
    In addition, I am attaching the waveform observed by outputting the AFE signal from the test terminal of PGA460.
    The time measurement synchronized with this waveform was 770μs.
    I would be happy if you could give me your opinion again. 

    Best, Ichi

    Data set in EEPROM REGISTER.pdfThreshold setting data.pdfwave_sample.pdf

  • 0 in reply to Ichi
    TI__Mastermind 21020 points

    Ichi-san,

    Isaac will be out of office for some time due to the holidays and E2E responses will be delayed.  We thank you for your patience until he returns and is able to take a closer look at your register settings.

    Thank you,

    Scott

  • 0 in reply to Scott Bryson
    TI__Mastermind 23045 points

    Hello Ichi-san,

    Thank you for the info and patience on this.

    Your EEPROM for the most part looks okay the first change I would recommend would be for

    TVGAIN- you might not need 64dB to start off you might be able to turn the initial gain down in INIT_GAIN register 1B

    Current TVG settings:                                                                   Suggested setting:

      

    Since I am not sure what your echo data dump looks like it is hard to make a suggestion on particular settings.

    FREQUENCY register 1C= B8 instead of B9 if your transducer is 400kHz this setting would get you closer closer.

    - I don't think your threshold configuration might be able to use some improvements I would try something more along the lines of this:

    Current threshold setting:                                                            Suggested setting:

    Again knowing how your echo looks would be helpful to adjust your threshold so this is just a suggestion.

    Remember if you are having a hard time visualizing how your thresholds and tvg settings would look you can always use the PGA460 GUI to see what those settings would look like even though you may not have an EVM.

    Best,

    Isaac